Fig. 3

Tissue-specific integrative analysis indicates potential effect on kidney and relation with eGFR GWAS loci. a eFORGE analysis for top 1000 eGFR CpGs: the x axis indicates tissues/cell type samples used in the analysis; the y axis shows eFORGE enrichment (−log10 p-value) of the CpG set with DNase I hotspots for a range of tissue samples (significant samples in black). The highest ranked sample set (highest black points) shows the most significant enrichment is for kidney samples, which are highly ranked for the top 1000 CpGs associated with eGFR. b FORGE2 analysis for eGFR SNPs from GWAS catalog: the x axis indicates tissues/cell type samples used in the analysis; the y axis shows FORGE2 enrichment (−log10 p-value) of the SNP set with DNase I hotspots for a range of tissue samples (significant samples in black). The highest ranked sample set (highest black points) shows the most significant enrichment also is for kidney samples, which are highly ranked for the top 249 SNPs associated with eGFR (taken from the GWAS catalog, https://www.ebi.ac.uk/gwas/, downloaded 10 April 2020). c TF motif enrichment results for EA probes driving eFORGE tissue-specific enrichment signal: the x axis indicates TF motifs from TRANSFAC, JASPAR, Taipale/SELEX, and Uniprobe databases; the y axis shows eFORGE-TF enrichment (−log10 hypergeometric p-value) of the input DMP set with TF motifs overlapping open chromatin sites for fetal kidney samples. Enrichment values for each TF motif are colored according to BY FDR-corrected q-value. A number of TF motifs involved in kidney development overlap top EA probes including OSR1, OSR2, TBX1, and PAX2. d Aggregated eFORGE results for EA probes: the x axis indicates sets of the top ranked DMPs used in the analysis (each set contains 1000 DMPs); the y axis shows eFORGE enrichment (−log10 p-value) of each of the DMP sets with open chromatin sites for kidney (red) and other tissue samples (gray). The highest ranked probe set (set 1, left) shows the most significant enrichment for kidney samples, which remain highly ranked for probe sets 2–5, in decreasing order of study p-value